CCJS300 Study Guide Exam 2 External Rival Factors Variables related to external validity Interactive testing effects making your sample different Interaction of selection bias and the experimental variable Reactive effects of experimental variable Hawthorne effect Multiple treatment interferences Two Other Factors Post hoc error just because B comes after A doesn t mean A caused B Placebo effect mind over matter Experimental Designs True experimental designs random assignment to assure group equivalence Quasi experimental designs no random maybe matching Pre experimental designs no group equivalence Classic Experiments Include Equivalence of groups at the start through random assignment R on table Pretests and posttests Experimental and control groups Examples of Classic Experimental Provo and Silverlake 1971 72 Researched the effect of experimental community based treatment programs in Provo Utah and Silverlake LA Experimental and control Experimental and control groups for both incarcerated and probation juveniles in both cities Result reduction in crime for both probation and incarceration after community based treatment program Kansas City Preventative Patrol Experiment 1974 15 beat area were divided into three sections reactive patrol proactive patrol and usual patrol Result no difference in patrol type police have much more leeway than expected when dealing with crime Minneapolis Domestic Violence Experiment 1977 Police called for domestic violence would either o Arrest the suspect o Separate them for 8 hours o Advise and mediate the suspects Researchers checked in every two weeks for six months Result Those who weren t arrested recidivated those who were arrested did not Advantages and Disadvantages of Experiments Advantages The Big Advantage Control Quick inexpensive Manageable can be replicated Disadvantages as they d like to May be hard to recruit respondents Experimenter effects biases The Big Disadvantage artificiality people don t have as much control o Double blind procedure experimenter doesn t know what respondent is being treated counteracts biases Sampling and Survey Research Sampling How we get our data They key principle in sampling is Representatives Two Types of Sampling Probability Samples permit estimation of the likelihood of each element of the population being selected to be in the sample SRS Simple Random Sample o Follows the EPSEM Rule equal probability selection method o Advantage enables us to use familiar statistics o Disadvantage need a complete list sampling frame to generate sample may not have it Stratified Random Sample looking at a variable ahead of time o Proportionate o Disproportionate oversampling Cluster Samples used in field interviews o Entire population divided into groups and a random sample is o Used when random sampling would be too scattered and collected expensive Systematic Sample Every nth item from the list of sampling frame Multistage samples complex combinations of other probability sampling techniques Nonprobability Samples any sampling procedure that violates the EPSEM Rule Quota Sample a nonprobability stratified sample Haphazard or convenience sample o Standing on street corner picking subjects no systematic way Purposive or Judgmental Sample o Based on your knowledge of a population and the purpose of the study choose people to your advantage Snowball Samples o Used in exploratory studies hard to get respondents accumulates over time o Existing study subjects recruit subjects among acquaintances Deviant or Extreme case sample o Selecting cases that are unusual or special in some way outstanding success or outstanding failure Sequential Sample o Similar to snowball and purposive o Keep finding cases until you run out of money resources etc Sample size probability and Central Limit Theorem As long as you have about 30 cases or more the sampling distribution will be approximately normally distributed Survey Research Typical Questions for Surveys Behaviors Attitudes Beliefs Opinions General or specific questions about knowledge Characteristics Expectations Self Classification Policies law enforcement knowledge What is a survey Comparing them to experiments Experiments Surveys o Random assignment treatment and control groups o Everyone answers all the survey questions o Some stimuli o Correlational Research Surveys Steps in Conducting a Survey 1 LOOK UP IN BOOK The Survey as Self Reported Data Disadvantages Potential Problems o Inaccurate reporting o Poor use of survey instruments o Poor research designs Advantages strengths of self reported surveys o Reliability and validity have improved o Check against other data o Use Polygraph Guidelines for Questionnaire Construction Question wording do it carefully o Variables list a list of variables being measured keyed to the question number in the questionnaire which are designed to measure each variable o Dummy variables preliminary blank tables constructed prior to data gathering that suggest the type of data needed as well as the type of data analysis Organization of the questionnaire o Put interesting non threatening questions first Types of Surveys 1 Mail Surveys o Advantages Wide geographic coverage Avoids interviewer bias Facilitates respondent privacy More time for responding o Disadvantages Non Response How to deal with non response Follow up postcard Offer incentives Altruistic beg Sponsorship endorsements Shortened format 2 Internet Surveys o Advantages Popular fast inexpensive o Disadvantages less educated 3 In person Interviews o Types Not a random sample May undercount females minorities the elderly the Structured close scheduled not as detailed similar to paper surveys Unstructured Depth very detailed takes longer so fewer respondents o General Rules Training and orientation Appearance demeanor of interviewer similar to the people Familiarity with questions Follow question wording exactly Record responses exactly Probe respondents for more complete answers and information o Advantages Personal contact between researcher and respondent Higher response rates than mail surveys Use audiovisual aids Clarify questions when needed Interviewer also an observer o Disadvantages Time consuming especially interviews done in the field Cost depending on needed sample size Interviewer effects or biases gender race or ethnicity 4 Telephone Surveys o Random Digit Dialing RDD random samples o Issues If respondent not available call backs may be used Prepare for distractions Include
View Full Document